Change Behavior in Pathfinder

Posted: May 16th, 2018 in News

As of Pathfinder 2018.2, Behaviors can now dynamically change from one to another throughout the simulation. The new Change Behavior step allows the user to define a progression to one of the other Behaviors defined in the model. This allows the user to alter the Occupants’ objectives after a specified time, after entering a certain room, or after other specific events.

Behaviors, as described in the Pathfinder [Download not found], “represent a sequence of actions the Occupant will take throughout the simulation.” These actions include steps such as Goto Waypoint, Wait, Assist Occupants, and Goto Exits. Change Behavior is a new type of step to randomly assign the Occupant a new Behavior based on a specified distribution. The Occupant will immediately begin the next sequence of actions once the change has executed.

Change Behavior
This list of behaviors models a scenario of occupants traveling from one room to another and eventually leaving the facility.

The Change Behavior step can randomly distribute a subsequent Behavior, including the current one, to each Occupant. In this example, we can have many behaviors that all lead to one another and one terminal behavior which tells the Occupant to leave the model. This scenario represents a shopping mall where Occupants may wander from one shop to the next through a central lobby, eventually leaving.

In this shopping mall, Occupants decide the next store they would like to visit upon exiting the current one.

Each Occupant walks into a room defined by their current behavior, waits between 10 to 30 seconds (to simulate the shopping experience), then the Change Behavior step selects their next Behavior. The distribution of behaviors selected within the Change Behavior step will eventually cause each Occupant to assume the Goto Any Exit Behavior.

Each Behavior is assigned a probability of selection.

There are many other ways to use the Change Behavior step in various applications. It can act as a filter to allow some Occupants to continue the second stage of a behavior while allowing others to switch to a different one. Perhaps this could model some occupants waiting on a rail car while others exit? We hope that you find creative uses for this feature and we are interested to see what you come up with.